diff --git a/litellm/main.py b/litellm/main.py
index b0215357e72..bead3e1d86b 100644
--- a/litellm/main.py
+++ b/litellm/main.py
@@ -298,7 +298,7 @@ def completion(
           "total_tokens": prompt_tokens + completion_tokens
         }
       response = model_response
-    elif hugging_face == True or custom_llm_provider == "huggingface":
+    elif custom_llm_provider == "huggingface":
       import requests
       API_URL = f"https://api-inference.huggingface.co/models/{model}"
       HF_TOKEN = get_secret("HF_TOKEN")
diff --git a/litellm/tests/test_secrets.py b/litellm/tests/test_secrets.py
index 2a4141bd49d..72e1bfb0860 100644
--- a/litellm/tests/test_secrets.py
+++ b/litellm/tests/test_secrets.py
@@ -19,7 +19,7 @@ messages = [{ "content": user_message,"role": "user"}]
 
 def test_completion_azure():
     try:
-        response = completion(model="gpt-3.5-turbo", deployment_id="chatgpt-test", messages=messages, azure=True)
+        response = completion(model="gpt-3.5-turbo", deployment_id="chatgpt-test", messages=messages, custom_llm_provider="azure")
         # Add any assertions here to check the response
         print(response)
     except Exception as e: